cloud-fan commented on a change in pull request #35248:
URL: https://github.com/apache/spark/pull/35248#discussion_r803289518
##########
File path: sql/core/src/test/scala/org/apache/spark/sql/jdbc/JDBCV2Suite.scala
##########
@@ -806,17 +806,51 @@ class JDBCV2Suite extends QueryTest with
SharedSparkSession with ExplainSuiteHel
checkAnswer(query, Seq(Row(29000.0)))
}
- test("scan with aggregate push-down: SUM(CASE WHEN) with group by") {
- val df =
- sql("SELECT SUM(CASE WHEN SALARY > 0 THEN 1 ELSE 0 END) FROM
h2.test.employee GROUP BY DEPT")
- checkAggregateRemoved(df, false)
+ test("scan with aggregate push-down: aggregate function with CASE WHEN") {
+ val df = sql(
+ """
+ |SELECT
+ | COUNT(CASE WHEN SALARY > 8000 AND SALARY < 10000 THEN SALARY ELSE 0
END),
+ | COUNT(CASE WHEN SALARY > 8000 AND SALARY <= 13000 THEN SALARY ELSE
0 END),
+ | COUNT(CASE WHEN SALARY > 11000 OR SALARY < 10000 THEN SALARY ELSE 0
END),
+ | COUNT(CASE WHEN SALARY >= 12000 OR SALARY < 9000 THEN SALARY ELSE 0
END),
+ | COUNT(CASE WHEN SALARY >= 12000 OR NOT(SALARY >= 9000) THEN SALARY
ELSE 0 END),
+ | MAX(CASE WHEN NOT(SALARY > 8000) AND SALARY >= 8000 THEN SALARY
ELSE 0 END),
+ | MAX(CASE WHEN NOT(SALARY > 8000) OR SALARY > 8000 THEN SALARY ELSE
0 END),
+ | MAX(CASE WHEN NOT(SALARY > 8000) AND NOT(SALARY < 8000) THEN SALARY
ELSE 0 END),
+ | MAX(CASE WHEN NOT(SALARY != 0) OR NOT(SALARY < 8000) THEN SALARY
ELSE 0 END),
+ | MAX(CASE WHEN NOT(SALARY > 8000 AND SALARY > 8000) THEN 0 ELSE
SALARY END),
+ | MIN(CASE WHEN NOT(SALARY > 8000 OR SALARY IS NULL) THEN SALARY ELSE
0 END),
+ | SUM(CASE WHEN NOT(SALARY > 8000 AND SALARY IS NOT NULL) THEN SALARY
ELSE 0 END),
+ | SUM(CASE WHEN SALARY > 10000 THEN 2 WHEN SALARY > 8000 THEN 1 END),
+ | AVG(CASE WHEN NOT(SALARY > 8000 OR SALARY IS NOT NULL) THEN SALARY
ELSE 0 END)
+ |FROM h2.test.employee GROUP BY DEPT
+ """.stripMargin)
+ checkAggregateRemoved(df)
df.queryExecution.optimizedPlan.collect {
case _: DataSourceV2ScanRelation =>
val expected_plan_fragment =
- "PushedFilters: [], "
+ "PushedAggregates: [COUNT(CASE WHEN ((SALARY) > (8000.00)) AND
((SALARY) < (10000.00))" +
+ " THEN SALARY ELSE 0.00 END), C..., " +
+ "PushedFilters: [], " +
+ "PushedGroupByColumns: [DEPT]"
checkKeywordsExistsInExplain(df, expected_plan_fragment)
}
- checkAnswer(df, Seq(Row(1), Row(2), Row(2)))
+ checkAnswer(df, Seq(Row(1, 1, 1, 1, 1, 0d, 12000d, 0d, 12000d, 12000d, 0d,
0d, 2, 0d),
+ Row(2, 2, 2, 2, 2, 0d, 10000d, 0d, 10000d, 10000d, 0d, 0d, 2, 0d),
+ Row(2, 2, 2, 2, 2, 0d, 12000d, 0d, 12000d, 12000d, 0d, 0d, 3, 0d)))
+ }
+
+ test("scan with aggregate push-down: aggregate function with NVL") {
Review comment:
can we test with UDF? something that we will never support SQL
generation.
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
To unsubscribe, e-mail: [email protected]
For queries about this service, please contact Infrastructure at:
[email protected]
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
